Evolution of the Social Brain

Both humans and dogs excel through their social abilities. These skills have also been argued to be the driving factor behind the astounding expansion of the neocortex in human phylogeny (Dunbar, 1998). The investigation of non-human primates cannot provide information on questions regarding convergent evolution (Fitch, Huber & Bugnyar, 2010) – e.g. whether the challenge to live in closely cooperating groups has (also) shaped our social skills. Domestic dogs (Canis familiaris) have thus been introduced as another target of comparative research. In this project funded by the WWTF, we will investigate the essential pillars of social cognition, such as action understanding, empathy, and theory of mind, in dogs as compared to humans. This project is an interdisciplinary collaboration and combines a variety of methods, such as behavioural measurements, eyetracking, and fMRI in awake and unrestrained dogs. 

Dogs show striking social skills, which have been suggested to be functional matches of the same skills in humans. Recent behavioural research has shown that dogs are able to discriminate facial emotion displays of humans (Müller et al, Curr. Biol. 2016). However, it is unclear whether this entails similar brain network as those engaged when humans recognize an emotion. We are investigating how dogs process facial expressions of happiness and anger as two examples of emotions often shown by humans when interacting with their dogs. To this end, dogs specifically trained to lie still in a human MRI scanner are watching videos of human demonstrators showing the respective emotions (see pictures to the right). First results were recently presented at the Canine Science Forum in Budapest. They suggest engagement of higher-order visual cortex areas as well as “limbic” structures, suggesting similar processing streams as those recruited in humans.

Neural bases of dog-human attachment

Humans and dogs share a long period of close interaction and co-habitation, shaping each other’s socio-affective behavior. Several studies have shown that the human-dog relationship resembles the human parent-child bond. Behavioral evidence suggests that dogs fulfill all four human attachment-defining criteria (proximity maintenance, separation distress, secure base, safe haven). But are these behavioral similarities based on the same underlying neuronal mechanisms? Are they comparable to those of the human child-mother bond? In this project, we investigate whether these networks are engaged in a similar fashion when dogs perceive their human attachment figure. Our main hypothesis is that “limbic” areas involved in attachment and affiliation, such as the meso-limbic reward pathway, show stronger activation in response to positive displays, and that this is linked to behaviour in the Strange Situation test for dogs (which assesses how dogs react to the presence or absence of their owners).

MRI scanning of fully awake and unrestrained dogs is a relatively new method used only by four pioneering labs world-wide, in Hungary, USA (Emory and Auburn) and Mexico. It has so far mainly used standard hard- and software developed for human MRI and fMRI. One of the aims of the CNU is to advance methodology for dog MRI. This includes development of a tailor-made dog head coil (together with the Radio Frequency Lab of the Medical University of Vienna ), of integrating eye tracking and behavioural data into dog fMRI scanning and analysis, and the development of tailored analysis approaches for direct quantitative comparisons between humans and dogs . Finally, we are also exploring novel dog training methods, and test their efficacy (see Training).